Estimation of path attenuation and site characteristics in the

north-west Himalaya and its adjoining area within Indian

Territory using generalized inversion method

Harinarayan, NH1, Abhishek Kumar

*2

1, 2 Department of Civil Engineering, Indian Institute of Technology Guwahati, Assam, India.

Corresponding to: Kumar Abhishek (abhitoaashu@gmail.com/ abhiak@iitg.ernet.in)

Abstract. Present work focuses on the determination of path attenuation and site characteristics of earthquake 1

recording stations, located in the north-west Himalaya and its adjoining region, within India. The work is done 2

using two-step generalized inversion technique. In the first step of inversion, non-parametric attenuation curves 3

are developed. ( ) ( ) as S wave quality factor within 105km, is obtained indicating that 4

the region is possibly heterogeneous as well as seismically active. In addition, presence of a kink is observed at 5

around 105km hypocentral distance while correlating path attenuation with the hypocentral distance, indicating 6

the presence of Moho discontinuity in the region. In the second step of inversion, site amplification curves are 7

developed separately from the attenuation corrected data for horizontal and vertical components of the 8

accelerograms. The amplification function and predominant frequency of each recording station based on 9

generalized inversion method is estimated and compared with that obtained from horizontal to vertical spectral 10

ratio (of S wave portion of the accelerograms) method. Path attenuation and site characteristics obtained in the 11

present study are very essential for developing regional ground motion model and for the seismic hazard 12

assessment of the above study area. 13

Keywords: Seismicity, Northwest Himalaya, path attenuation, site characteristics, Generalized Inversion, HVSR 14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

1 Introduction 32

The Himalayan arc extending approximately 2500km between Kashmir and Arunachal Pradesh is one among 33

the seismically most active regions across the globe. Seismic activity of this region can be understood based on 34

induced damages witnessed primarily during 4 great earthquakes (EQs) including 1897 Shillong EQ, 1905 35

Kangra EQ, 1950 Assam EQ and 1934 Bihar-Nepal EQ, occurred in the last 120 years. Based on seismic 36

activity, the entire Himalayan belt can be subdivided into three distinct segments namely the western, the central 37

and the eastern Himalayas (Philip, 2014). The region of the north-west Himalaya and its foothills, within India 38

come under seismic zone IV and V as per IS 1893: 2016, indicating regions of high to very high seismicity. 39

Intensity of ground shaking during an EQ, at a particular site is a collective effect of source, path and 40

site parameters. Source parameters include magnitude, fault mechanism, stress drop and rupture process. On the 41

other hand, path parameters include geometric attenuation and loss of seismic energy due to the anelasticity of 42

the earth and scattering of elastic waves in heterogeneous media. Similarly, site characteristics include 43

modification of amplitude, frequency content and duration of the incoming seismic wave by subsurface medium 44

as reaches the surface. Determinations of aforementioned EQ parameters are important for the development of 45

region specific ground motion models, which can further be used for region/site specific seismic hazard 46

assessment (Baro et al., 2018). Above parameters can be estimated from EQ records based on some spectral 47

modelling or inversion approach like generalized inversion method (Andrews, 1986; Castro et al., 1990; Oth et 48

al., 2009 etc.). 49

To understand the on-going seismicity of various regions within India, the Government of India has 50

installed number of EQ recording stations. EQ records from these stations since 2004 are maintained by 51

PESMOS (Program for Excellence in Strong Motion Studies), which is currently one of the most significant 52

resource of ground motion records in India. At present, PESMOS manages EQ records from 300 recording 53

stations which are distributed in the northern and northeaster parts of India as well as in the Andaman and 54

Nicobar Islands (Kumar et al. 2012). It must be highlighted here that PESMOS database is lacking in terms of 55

accurate information about subsurface for majority of recording stations (Harinarayan and Kumar, 2018). Site 56

class given by PESMOS is based on physical description of surface materials, local geology following 57

Seismotectonic Atlas of India (GIS 2000) and Geological Maps of Indian (GSI, 1998) and not based on actual 58

field investigation (Kumar et al., 2012). In the absence of accurate information of local soil, utilizing EQ records 59

from PESMOS database for seismic studies is a major challenge. Geophysical subsurface exploration studies on 60

some of the recording stations in northwest Himalaya reported by Pandey et al., (2016a; b) had highlighted the 61

flaws in site class given by PESMOS. There are recording stations classified to be on rock site but were found to 62

be on soil sites by Pandey et al., (2016a; b). Harinarayan and Kumar (2018) also reported problems in the 63

subsurface information given by PESMOS and attempted site classification of PESMOS recording stations 64

In the present study, EQ records from the region of the north-west Himalaya and its foothills within 65

India, as obtained from PESMOS database, are analysed for estimating path attenuation and site characteristics 66

separately, using a two-step generalized inversion of the S-wave Fourier spectra (hereafter referred to as GINV). 67

In the first step, attenuation curves are developed using a non-parametric inversion approach (similar to Castro 68

et al., 1990 and Oth et al., 2008). In the conventional generalized inversion method (Andrews, 1986; Hartzell, 69

1992), the second step of inversion calculates both site and source spectra, by inverting the S-wave (or Coda 70

wave) spectra, corrected for the path parameter. This method however requires one or more reference sites 71

(usually rock sites) in-order to remove the trade-off between the source and site parameters (Andrews 1986). In 72

the absence of subsoil information for majority of recording stations managed by PESMOS as highlighted 73

above, identifying reference site is not possible. For this reason, only the site parameters are evaluated in the 74

second step of inversion, using a non-reference generalized inversion approach (similar to the work by Joshi et 75

al., 2010; Harinarayan and Kumar 2017b). Further, Obtained site terms are compared with the one calculated 76

from horizontal to vertical spectral ratios (HVSR) from the same S-wave window as used in the GINV. 77

This study is one of its kind, which systematically evaluates path and site parameters using a larger and 78

regional database. Existing studies on attenuation characteristics determination for the northwest Himalaya used 79

few EQ records that too from limited recording stations. Joshi (2006) estimated frequency independent S wave 80

quality factor ( ) for the Garhwal Himalayas using 1991 Uttarkashi EQ and 1999 Chamoli EQ ground motion 81

records from 8 recording stations. In another study, Singh et al., (2012) estimated frequency depended for the 82

Kumaun Himalayas using 23 EQ events, from 9 recording stations by applying the extended coda-normalization 83

method. Similarly, Negi et al., (2015), Banerjee and Kumar, (2017) and Tripathi et al., (2015) estimated for 84

the Garhwal Himalayas. The aforementioned studies did not highlight the attenuation characteristics of the 85

entire north-west Himalaya region. 86

Similar to path attenuation studies, very few studies on the determination of site characteristics from 87

EQ records also exist for this region. Nath et al., (2002) computed site terms using the aftershocks of the 1999 88

Chamoli EQ, obtained from 5 recording stations located in the Uttarakhand region. Similarly, Sharma et al., 89

(2014) estimated site parameters for the Garhwal region of Uttarakhand using EQ records in context of 90

generalized inversion and HVSR. In another work, Harinarayan and Kumar (2017) reported a comparative study 91

on site characteristics computed using EQ records from Tarai region of Uttarakhand using multiple analytical 92

approaches. In another recent work, Harinarayan and Kumar (2018) computed site parameters for recording 93

stations in the northwest Himalayas in terms of predominant frequency (fpeak) alone using HVSR method. 94

2 Study Area 95

Present study area includes states of Himachal Pradesh, Uttarakhand, Punjab, Haryana and national capital city, 96

New Delhi, covering an area between 28º N to 34º N latitude and 75.8º E to 80.5º E longitude. According to 97

2011 Census, the region has a population of 96 million. From seismicity point of view, major Seismotectonic 98

features of the present study are characterized by three north-dipping thrust systems such as the Central thrust 99

(MCT), the Main Boundary thrust (MBT) and the Himalayan frontal thrust (HFT) (Valdiya, 1981). Other 100

tectonic features includes, the Jhelum Balakot fault, the Drang thrust, the Lesser Himalayan Crystalline Nappes, 101

the Jammu thrust, the Vaikrita thrust, the Karakoram fault, the Jwala Mukhi thrust, and the Ramgarh thrust. 102

Both the MCT and the MBT lie parallel to each other within western Himalayan region and were produced 103

during the Cenozonic shortening (Malik and Nakata 2003). The HFT is the youngest active thrust separating the 104

Himalaya region and the Indo-Gangetic alluvial plain (Kumar et al. 2009). The HFT, the MBT and the MCT 105

have generated major EQs in this region (Philip et al. 2014). 106

Two of the most damage inducing EQs, in the last 120 years, in the west Himalayan regions include 107

1905 Kangra-Himachal Pradesh EQ (Ms=7.8) (Ambraseys and Douglas 2004) and 2005 Muzzafarbad-Kashmir 108

EQ (Mw=7.6) (Avouac et al. 2006). Both of these EQs caused severe loss to life and property. Recent EQs of 109

1991 Uttarkashi (Mw=6.8) and 1999 Chamoli (Mw=6.6) had occurred on the MCT zone (Harbindu et al. 2014). 110

The 1999 Chamoli EQ caused a huge landslide in Gopeshwar situated less than 2km northwest of Chamoli city 111

(Sarkar et al. 2001). This EQ also caused shaking in Chandigarh and Delhi, located far away from the epicentre 112

(Mundepi et al., 2010). 113

3 Database 114

Ground motion records used in this study consists of three components accelerograms obtained from PESMOS 115

database available at http://www.pesmos.in/. The instrumentation used for recording EQs consists of internal 116

AC-63 GeoSIG triaxial force balanced accelerometers and GSR-18 GeoSIG 18 bit digitizers with external GPS 117

(Kumar et al., 2012). Further, ground motion recordings are done in trigger mode during each EQ with a 118

sampling rate of 200 per second. 119

For the present analysis, ground motion records of EQs happened between 2004 and 2017, available on 120

PESMOS are used. For estimating site characteristics, 341 records from 86 EQs, with magnitudes ranging from 121

Mw=2.3 to 5.8, having focal depths ranging from 2 to 80km are used. Further, these records are corresponding 122

to 101 recording stations, located in the hypocentral distance ranging from 9 to 355km. Coordinates of each of 123

the recording station, used in this work are listed in Table 1, columns 2 and 3. Further, details of EQs used for 124

estimating site characteristics are summarized in Table 2. 125

For estimating path attenuation however, only those EQs recorded at atleast two recording stations 126

among which at least one recording station is located within hypocentral distance equal to or less than the 127

reference distance (reference distance is discussed under section ‗Spectral attenuation with distance) are 128

considered. Out of 341 EQ records used for estimating site parameters, only 207 EQ records satisfies the above 129

mentioned reference distance criteria. Final database for estimating path attenuation consists of 207 records 130

from 32 EQs, recorded at 69 recording stations, with magnitude (Mw) in the range of 3.1 to 5.5, focal depths 131

from 3 to 55km and hypocentral distance from 9 to 200km. Table 3 summarizes the details of the dataset used 132

for estimating path attenuation. In addition, Figure 1 shows the source-to-recording station distance of the data 133

set used in the present study. 134

3.1 Data processing 135

Signal to pre-event noise (all of equal window length) ratio (SNR) for all the records are computed and records 136

with SNR≥5 (similar to Ameri et al., 2011) are considered for further analyses. All the EQ records are corrected 137

for baseline correction following a 5% cosine taper and a band-pass filtering, between the frequency range of 138

0.25Hz and 15Hz, using a Butterworth filter. Further, time windows starting about 0.5s before the onset of the S 139

wave and ending when 90% of the total seismic energy of the EQ record is reached, are separated and tapered 140

with a 5% cosine window (Ameri et al., 2011; Bindi et al., 2009). Typical lengths of the time windows for the 141

present analysis vary from 4 to 15s. Further, for some of the records, where the window lengths obtained to be 142

longer than 15s, are fixed to 15s in order to minimize coda wave energy in the analysing time window (Oth et 143

al., 2008). Later, based on the extracted windows, the Fourier amplitude spectra is calculated for each EQ 144

record, smoothened by applying the Konno and Ohmachi (1999) algorithm, with the smoothing parameter ―b‖ = 145

20. 146

For further analyses, path and site parameters are estimated using above processed EQ records, based on two 147

separate inversion procedures as discussed separately in the following sections. 148

4 Path attenuation 149

In the first step of inversion, path attenuation curves are developed by eliminating the effect of site parameter, 150

thereby retaining only the source and path attenuation characteristics. All EQ records, irrespective of whether 151

located on soil or rock sites can be utilized for inversion. This way, present method is very much suitable for 152

PESMOS database where accurate subsurface information of recording stations is not available. The horizontal 153

portion of the accelerograms (obtained by the root mean square average of the east-west and north-south 154

components) is considered for developing path attenuation curves. Detailed discussions on the method can be 155

found in following sub-section. 156

4.1 Methodology 157

Following Castro et al., (1990), observed spectral amplitude (acceleration) ( ), of EQ , at recording 158

station , and frequency f can be modelled linearly as: 159

( ) ( ) ( ) (1) 160

Here, ( ) is a scalar, which is governed by the magnitude of the EQ (one value for each EQ). ( ) is the 161

attenuation function and is independent of the magnitude of EQ. Here, ( ) incorporates both geometric 162

spreading and anelastic attenuation variation with the hypocentral distance. It has to be mentioned here that 163

( ) in Eq. (1) is not limited to a particular functional form, instead, is assumed to decay smoothly with 164

hypocentral distance ( ) and thus take the value of unity at a reference distance ( ), as given in Eq. 2 (Castro 165

et al., 1990; 1996; 2003). 166

( ) (2) 167

Model given by Eq. (1) has no factor representing site effect and is contained in both ( ) 168

and ( ). Any rapid undulations in ( ) are due to the absorbed site effects (Oth et al., 2008). Two 169

weighing factors, and are incorporated in the Eq. (1) following Castro et al., (1990). is used to 170

smoothen the attenuation term with distance curve by supressing the undulations and thereby removing any 171

absorbed site effects from ( ). is used to impose ( ) constraint, as mentioned earlier. The 172

value of and here is chosen reasonably such that the site effects are supressed but the change in the 173

attenuation characteristics with distance can be observed (Oth et al., 2008). Solution to Eq. (1) in the matrix 174

form, after incorporating weighing factors, is obtained using singular value decomposition method, (discussed in 175

detail in Appendix A). 176

4.2 Spectral attenuation with distance 177

Figure 2 shows the number of EQ records for various hypocentral distance range considered. It can be observed 178

from Figure 2 that there are very less EQ records available beyond hypocentral distance of 115km. For this 179

reason, EQ records with hypocentral distance up to 115km are only considered for the determination of path 180

attenuation. The constraint ( ) is applied at =15km, irrespective of the frequency. The hypocentral 181

distance range from 15 to 115km is divided into 10 bins, each bin having 10km width. Further, attenuation 182

curves are computed for each of the selected 17 frequencies from 1Hz to 15Hz (see Table 4, column 1). 183

Variation of attenuation curves with hypocentral distance, obtained in the present study, for the selected 184

frequencies is depicted in Figure 3. Based on Figure 3, a general trend in which attenuation curves exhibit decay 185

with hypocentral distance up to 105km can be observed. Beyond 105km a kink is observed as seen in Figure 3. 186

This kink in the attenuation curves beyond 105km is very distinct and clear at lower frequencies (<5.5 Hz). 187

Bindi et al., (2004) and Oth et al., (2011) reported a similar trend in the attenuation curves for the Umbria 188

Marche and Japan regions respectively. Oth et al., (2011) attributed this behaviour to the combined effect of 189

reflected or refracted wave arrivals from the Moho in Japan. Presence of Moho in the North-west Himalaya was 190

reported by Saikia et al., (2016) based on Teleseismic receiver function analysis. Referring to Oth et al., (2011) 191

work, presence of kink beyond 105km in attenuation curves, as obtained in this study may also be due reflected 192

or refracted waves from the Moho. Further, detailed study in this direction can be done in the future and is 193

beyond the presence scope of the work. Observing the attenuation curves at different frequencies as given in 194

Figure 3, it can concluded that attenuation curves at higher frequencies (>5.5Hz) decay more rapidly compared 195

to lower frequencies for the present study region. This observation is consistent with the findings by Castro et 196

al., (2003) for Guadeloupe (France) and Oth et al., (2011) for Japan. 197

Further, for the kink observed at 105km, in case of lower frequencies, its sharpness reduces with 198

increasing frequency, as can be observed in Figure 3. At frequencies greater than 10Hz, the kink at 105km 199

smoothen and the attenuation curves beyond 105km for frequencies greater than 10 Hz becomes flat as observed 200

in Figure 3. This change in the characteristics of the kink at higher frequency indicates that the arrival of waves 201

from the Moho also get attenuated more at higher frequencies in comparison to lower frequencies. 202

4.3 Quality factor ( ) estimation 203

In order to estimate , inversion is repeated, however only considering records within hypocentral distance in 204

the range 15km to 105km, where a monotonic decrease in attenuation curves with hypocentral distance is 205

observed (see Figure 3). The attenuation curves are modelled in terms of geometric spreading ( ) and 206

quality factor ( ) in accordance with Castro et al., (1996) as; 207

( ) ( ) [

] (3) 208

Where, is the frequency and is the mean shear wave velocity in the crustal medium taken as 3.5km/s as per 209

Mukhopadhyay and Kayal, (2003). Further, ( ) is considered as in accordance with Banerjee and 210

Kumar (2015) for this region. For each frequency considered in this study (see Table 4), Eq. (3) is linearized by 211

taking logarithm and corrected for the effect of ( ) as given in Eq. (4). 212

( ) ( )

(4) 213

Ascribed to Castro et al., (2003), Eq. 4 is written in the form; 214

( ) (5) 215

Where ( ) and are given as; 216

( ) ( ) ( ) (6) 217

(7) 218

Where, m in Eq. 5 represents the slope between ( ) and based on a linear least-square fit, obtained for each 219

of the selected frequencies. Further, the values are estimated for the selected frequencies by substituting the 220

value of computed using Eq. (7). Columns 2 and 3, Table 4 list the value of and with frequency ( ) 221

respectively. In order to build the frequency dependent relationship ( ), the value of is fitted as a 222

function of frequency using a power law. In the above expression, is the frequency dependent coefficient, 223

which is approximately equal to 1 and varies on the basis of the heterogeneity of the medium (Aki 1980). 224

Variation of against frequency as illustrated in Figure 4 gives frequency dependent for the north-west 225

Himalayas as; 226

( ) ( ) (8) 227

Values of and (in the expression ) are attributed to the level of tectonic activity and degree of 228

heterogeneity respectively, present in the region. Aki (1980) concluded higher values of for tectonically active 229

regions in comparison to that of stable regions. Similarly, low value of (<200) is an indication of larger 230

degree of heterogeneities in the medium (Joshi 2006). The values of (=0.94) and (=105), obtained in this 231

study indicate that the present study region is tectonically active, characterized by higher degree of 232

heterogeneities, in accordance with Aki (1980) and Joshi (2006). 233

4.4 Comparison with regional and global attenuation characteristics 234

As discussed earlier, numerous studies exist where path attenuation of different parts of the present study area 235

were attempted in the past. Comparison of present results with those obtained by the previous researchers for the 236

northwest Himalaya and Delhi region are attempted as shown in Figure 5. It can be seen from Figure 5 that the 237

attenuation curve obtained in the present study falls in between existing attenuation curves for the different parts 238

of the north-west Himalaya as given in the literature, [Kinnaur, (Kumar et al., 2009), Kumoan (Mukhopadhyay 239

et al., 2010), Garhwal regions (Negi et al., 2015) and Delhi (Sharma et al., 2015)]. It has to be highlighted here 240

that the database for the present study also includes EQ records from Kinnaur, Kumoan, Garhwal regions of the 241

north-west Himalayas as well as from regions in and around Delhi. For this reason, the value of and 242

obtained in the present study reflects an average attenuation characteristics of regions encompassing north-west 243

Himalaya up to Delhi region but within Indian boundary. 244

Furthermore, the attenuation results obtained in this study is compared with some typical results 245

obtained globally in terms of attenuation characteristics and tectonic setting as shown in Figure 6. Literature 246

suggests low values of for tectonically active regions [e.g. Kato Japan region (Yoshimoto et al., 1993); East 247

central Iran (Mahood et al., 2009); Egypt (Abdel 2009); and Umbria–Marche region (Lorenzo et al., 2013)]. 248

Similarly, relatively high values of were reported for tectonically stable areas [e.g. Baltic Shield (Kvamme 249

and Havskov 1989); Central South Korea (Kim et al., 2004) and South Eastern Korea (Chung and Sato 2001)]. 250

Attenuation values obtained in the present study show good agreement with other studies having lower 251

across the globe. Further, attenuation curves for the present region is also found closer to regions of high 252

seismicity like Umbria–Marche and Eastern Iran as can be observed from Figure 6. 253

5 Site Effects 254

After estimation of path parameter as discussed in the previous section, site characteristics of the recording 255

stations are determined in the second step of inversion. In addition to GINV, site components are also estimated 256

using HVSR method. Detailed discussion on GINV and HVSR is given in the subsequent sections. 257

5.1 GINV 258

GINV was developed by Andrews (1986) by improvising spectral ratio method. Since then, various forms of 259

this technique have been developed and used for estimating the seismic site characteristics by various 260

researchers (Castro et al., 1990; Boatwright et al., 1991; Oth et al., 2008 etc.). The methodology used for 261

estimating site characteristics in the present study is discussed here. 262

As per Iwata and Irikura (1988), the Fourier amplitude (acceleration) spectrum (FAS) of the EQ 263

recorded, at the recording station, ( ) can be represented in the frequency domain as the product of 264

source term ( ( ) ), path attenuation ( ( ) ) and site term ( ( ) ) as shown below; 265

( ) ( ) ( ) ( ) (9) 266

Further, the path attenuation term can be removed from the spectral content of the record following Andrews 267

(1986) as; 268

( ) ( )

( ) ( ) ( ) (10) 269

The value of ( ) can be estimated using Eq. (3) and by considering as per Eq. (8), obtained in the earlier 270

section. Further, Eq. (10) can be linearized, by taking natural logarithms on both sides as per Andrews (1986) 271

giving; 272

( ) ( ) ( ) (11) 273

Considering: ( ), ( ) ( ) and ( ) j, Eq. (11) in the matrix form can be 274

written in accordance with Joshi et al., (2010) and following the notations of Menke (1989) as; 275

← 1st event → ← nth event → ← Site effect→

1 2 … m

1 2 … m 1 2 … m

1 0 … 0 …... 0 0

0 1 0 … 0

s1(f1)

d1(f1)

0 1

0 …... 0 0

0 0 1 … 0

:

:

:

:

: :

: : :

:

:

:

:

:

: :

: : :

:

:

:

0 0 0 1

0 0 … 0 0 0 … 1

s1(fn)

d1(fm)

sn(f1) =

:

For nth earthquake

sn(fn)

dn(fm)

0 0 … 0 ….. 1

… 0 1 0 … 0

g(f1)

dn(fm)

0 0 … 0 ….. 0 1 … 0 0 1 … 0

g(f2)

:

: :

:

: :

: : :

:

:

:

: :

:

: :

: : :

:

:

:

0 0 … 0 … 0 0 … 1 0 0 … 1

g(fm)

dn(fm)

(12) 276

The matrix form in Eq. (12) represents a purely under determinate system since there are ( ) 277

parameters for ‗ data (here is the number of sample frequency and is the number of EQs recorded at 278

a particular recording station). In here, Eq. (12) is solved using Moore- Penrose matrix inversion procedure 279

(minimum norm inversion) given by Penrose, (1955) to determine ( ) at each of the recording station. 280

Based on the above discussed methodology, inversions are performed for east-west, north-south and 281

vertical components of EQ records separately to obtain the amplification curves in the frequency range of 282

0.25Hz to 15Hz, for each of the three components. For further calculation, the horizontal component is obtained 283

as the geometric mean of east-west and north-south components. 284

5.2 HVSR 285

HVSR method is an extension of Nakamura (1989) technique, which is widely to assess the subsoil 286

characteristics using recorded ambient noises. Nakamura (1989) technique is based on the assumption that the 287

soil amplification effects are retained only in the horizontal component whereas the source and the path effects 288

are maintained both in vertical as well as horizontal components of ground motion. Hence, the ratio of 289

horizontal and vertical components gives an estimate of site amplification. Lermo and Chavez-Garccia (1993) 290

extended Nakamura (1989) technique to S wave part of the accelerograms and studied the theoretical basis of 291

the technique by numerical modelling oSV waves. Later, HVSR method was applied to EQ recordings 292

worldwide (Luzi et al., 2011; Yaghmaei-Sabegh and Tsang 2011; D‘Alessandro et al., 2012; Harinarayan and 293

Kumar 2017a, b, 2018 etc.) to obtain the site characteristics. 294

Comparative studies between HVSR and other methods of evaluating site parameters reported by Field and 295

Jacob (1995), Parolai et al., (2004), Shoji and Kamiyama (2002) Harinarayan and Kumar (2017b) etc. show that, 296

HVSR can provide good and reliable estimate of predominant frequency. However, the above literatures also 297

point out discrepancies in amplification levels obtained from HVSR with other methods. In order to compare the 298

site amplification functions obtained from HVSR and GINV methods, HVSR for each station is computed 299

considering the same S wave window as used in the GINV method. HVSR for each recording station is 300

determined using the following steps; 301

1. Calculate the response spectra considering 5% damping, for all the three components (north-south, east-302

west and vertical) of ground motion records. 303

2. Obtain the geometric mean of the two horizontal response spectra components (H) using Eq. (13) given 304

below; 305

( ) (13) 306

3. Calculate the ratio of to ( ). 307

Where, HEW and HNS are the response spectrum of the horizontal east-west and north-south components 308

respectively and V is the response spectrum corresponding to vertical component of ground motion. Then, the 309

HVSR at each of the recording station can be estimated as; 310

( ) ∑

(14) 311

Here, is the number of events recorded at recording station ― ‖ and ( ) indicates the average HVSR 312

value for a particular recording station ― ‖. The fpeak is the value of frequency corresponding to a maximum 313

value of (denoted by Apeak) at the recording station ― ‖. 314

5.3 Site Parameters 315

Site amplitude (SA) curves are developed using GINV for the horizontal (GINV H) and the vertical components 316

(GINV V). Figure 7 shows typical amplification curves obtained for GINV H (indicated by dashed lines) and 317

GINV V (indicated by firm lines) for typical 6 recording stations. In general, obtained amplification values for 318

GINV H is greater than GINV V for all frequencies. A typical observation (from Figure 7) for both GINV H and 319

GINV V is that the high level of amplification is observed at high frequencies. For several recording stations, 320

clear and distinct peak in the amplification curve can be observed (e.g. JAMI, BAR, GHA and SND) from 321

Figure 7. Moreover, the overall peaks of the GINV V component are usually at higher frequencies than for the 322

GINV H component. Further, in case of few recording stations, the shift in frequency is relatively close to a 323

factor √ (eg. BAR and GHA). Next, Site amplification factor (SAF) is estimated based on the GINV results 324

(denoted by GINV H/V) as the ratio of GINV H to GINV V. The value of frequency corresponding to the 325

maximum value of SAF (denoted as Apeak) is fpeak. GINV H/V curves are compared with those estimated using 326

HVSR method for a total of 101 recording stations. Figure 8 shows the comparison of the HVSR (indicated by 327

dashed lines) and GINV H/V (indicated by firm line) for typical 9 recording stations from above analyses. A 328

general observation made from Figure 8 is that both HVSR and GINV H/V show similar SAF patterns for all 329

recoding stations in the selected frequency range. Overall value of fpeak obtained exhibit 1:1 matching between 330

the two methods. However, there is trend of difference in terms of Apeak values. Apeak values obtained using 331

HVSR are found higher compared to those obtained using GINV H/V curves. This observation is also reported 332

in other regions (Sharma et al., 2014; Field and Jacob, 1995). The values of fpeak obtained using GINV H/V and 333

HVSR are tabulated in Column 5 and 7, Table 1 respectively. Similarly, the values of Apeak obtained using 334

GINV H/V and HVSR are tabulated in Column 6 and 8, Table 1 respectively. The maximum value of fpeak of 335

15Hz is observed for the recording station GGI with a value of Apeak of 5.3 based on GINV H/V. The maximum 336

value of Apeak of 12.2, based on GINV H/V is observed for ADIB recording station at 6.3Hz. The range of Apeak 337

based on HVSR varies between 1.7 and 19.4, while based on GINV H/V, the range of Apeak varies between 1.5 338

and 12.7. Similarly, of the fpeak based on HVSR varies between 0.4Hz and 10Hz, while based on GINV H/V, the 339

fpeak varies between 0.5Hz and 15Hz. 340

Further, based on the value of fpeak obtained above, using GINV, the recording stations are classified as 341

either rock sites or soil sites. In general, criteria based on average shear wave velocity over 30m ( ) is used 342

for site classification. A site can also be classified based on fpeak values. Such an approach was used by 343

Harinarayan and Kumar (2018) to classify the recording stations in the north-west Himalaya based on fpeak 344

obtained using HVSR method, where stations having fpeak less than 6.35Hz were classified as soil sites and 345

stations having fpeak greater than 6.35Hz were classified as rock sites. The range of fpeak values reported by 346

Harinarayan and Kumar (2018) for soil and rock sites were calculated based on the range of based on 347

NEHRP site classification scheme (BSSC, 2003) in accordance with the Eq. (15) (Kramer, 1996), correlating 348

to soil depth (denoted by and taken as 30m) and shear wave velocity ( ). 349

⁄ (15) 350

Based on NEHRP classification criteria (BSSC, 2003), all the recording stations in the present are classified as 351

either rock site or soil site as given in Column 9, Table 1. Out of 101 recording stations 10 recording stations are 352

classified as rock sites and the rest 91 recording stations are classified as soil sites. 353

5.5 Relationship between GINV H/V results and Vs30 354

The value of are available for 8 recording stations in Tarai region of Uttarakhand and 19 stations in Delhi 355

region from Pandey et al., (2016a) and Pandey et al., (2016b) respectively based on MASW test. These 27 356

recording stations coincides with recording stations where Apeak and fpeak are determined in the present study. 357

Vs30 (obtained from Pandey et al., 2016 a, b), fpeak and Apeak (as per present work) for above 27 recording stations 358

are listed in Table 5. Based on the present findings, relationship between fpeak and Vs30 for the 27 recording 359

stations is proposed as shown in Figure 9a having R2=0.71 as; 360

( )( ) ( ) (16) 361

Similarly, the relationship between Apeak and Vs30 for above 27 recording stations, as obtained in the present 362

study (see Figure 9b) is as follows; 363

( )( ) ( ) (17) 364

A lack of correlation (correlation coefficient =0.47) between and Apeak is observed as shown in Figure 9b, 365

which is also reported in the previous studies like Dutta et al., (2001), (2003) and Hassani et al., (2011). It can 366

be concluded from the proposed correlations (Eqs. 16 and 17) that the value of fpeak increases with increase in 367

Vs30 whereas the value of Apeak decreases with the increase in Vs30. It has to be highlighted here that both the 368

equations [Eqs. 16 and 17] are applicable for sites having fpeak in the range 1.8 to 6 Hz, and Apeak in the range 2 369

to 6.9. 370

Conclusion 371

In the light of on-going seismicity and catastrophic damages witnessed in the past, determination of path 372

attenuation as well as site characterization of EQ recording stations of PESMOS located in the north-west 373

Himalayas and adjoining area is attempted. EQ recorded between 2004 and 2017 are used in the analyses based 374

on two-step inversion. While determining path attenuation, a kink at 105km hypocentral distance is observed in 375

this work. Referring to similar observations from other regions, presence of Moho discontinuity is proposed in 376

the region. This finding can be validated based on detailed study and is beyond the scope of present work 377

objective. Further, based on attenuation curve obtained till 105km hypocentral distance and over wide range of 378

frequencies, ( ) ( ) is obtained for the present study area, clearly indicating that the 379

region is heterogeneous and seismically active. 380

In absence of proper geological information of PESMOS recording stations as highlighted by numerous 381

recent studies, site classification of all the recording stations considered in this study are done based on GINV 382

and HVSR. Based on this analysis, out of 101 recording stations, 10 are found to be located on rock while 91 383

stations are found to be located on soil sites. 1:1 matching for all the recording station from GINV and HVSR 384

further enhances the confidence on present findings. Based on the findings, two empirical correlations for Apeak 385

and fpeak with Vs30, are proposed. 386

Path attenuation and site characteristics are the key factors to be used for developing regional ground 387

motion model. Further, with EQ catalogue known, present findings can be used for detailed seismic hazard 388

assessment of the study area. In addition, identifying recording station whether considered recording stations are 389

located on rock sites or soil sites will help in utilizing ground motion records for scenario based seismic hazard 390

assessment. 391

392

393

Authors Contribution: 394

Harinarayan N H developed code generalized inversion, analyzed the records and all relevant literature review. 395

Kumar Abhishek (AK) highlighted the importance of site characterization for PESMOS recording stations and 396

need for the study. 397

Acknowledgement 398

The authors would like to thank the INSPIRE Faculty program by the Department of Science and Technology 399

(DST), Government of India for the funding project ‗‗Propagation path characterization and determination of in-400

situ slips along different active faults in the Shillong Plateau‘‘ ref. no. DST/INSPIRE/04/2014/002617 [IFA14-401

ENG-104] for providing necessary funding and motivation for the present study. 402

Reference 403

Abdel-Fattah, A. K.: Attenuation of body waves in the crust beneath the vicinity of Cairo Metropolitan area 404

(Egypt) using coda normalization method, Geophys. J. Int., 176(1), 126-134, 2009. 405

Aki, K.: Attenuation of shear-waves in the lithosphere for frequencies from 0.05 to 25 Hz, Phys. Earth Planet. 406

Inter., 21, 50–60, 1980. 407

D‘Alessandro, C., Bonilla, L.F., Boore, D.M., Rovelli, A., and Scotti, O.: Predominant-period site classification 408

for response spectra prediction equations in Italy, Bull. Seismol. Soc. Am., 102(2), 680–695, 2012. 409

Ameri, G., Oth, A., Pilz, M., Bindi, D., Parolai, S., Luzi, L. and Cultrera, G.: Separation of source and site 410

effects by generalized inversiontechnique using the aftershock recordings of the 2009 L‘Aquila earthquake, 411

Bull. of Earthquake Eng., 9(3), 717-739, 2011. 412

Andrews, D. J.: Objective determination of source parameters and similarity of earthquakes of different size, 413

Earthquake source mechanics, 259-267, 1986. 414

Banerjee, S., Kumar, A.: Determination of seismic wave attenuation for the Garhwal Himalayas, India. Geosci 415

Res 2(2):105–126, 2017. 416

Baro, O., Kumar, A., and Zadeh, A.S.: Seismic hazard assessment of the Shillong Plateau, India, Geomatics, 417

Nat. Hazards and Risk, 9:1, 841-861, 2018. 418

Bilham, R., Larson, K., and Freymueller, J.: GPS measurements of present-day convergence across the Nepal 419

Himalaya, Nature, 386(6620), 61, 1997. 420

Bindi, D., Castro, R. R., Franceschina, G., Luzi, L., and Pacor, F.: The 1997–1998 Umbria‐ Marche sequence 421

(central Italy): Source, path, and site effects estimated from strong motion data recorded in the epicentral area, J. 422

Geophys. Res. B: Solid Earth, 109(B4), 2004. 423

Bindi, D., Pacor, F., Luzi, L., Massa, M., and Ameri, G.: The M w 6.3, 2009 L'Aquila earthquake: source, path 424

and site effects from spectral analysis of strong motion data, Geophys. J. Int., 179(3), 1573-1579, 2009. 425

BSSC (2003). Building Seismic Safety Council, NEHRP recommended provisions for seismic regulations for 426

new buildings and other structures. Report FEMA-450 (Provisions), Federal Emergency Management Agency 427

(FEMA), Washington. 428

Boatwright, J., Fletcher, J. B., and Fumal, T. E.: A general inversion scheme for source, site, and propagation 429

characteristics using multiply recorded sets of moderate-sized earthquakes, Bull. Seismol. Soc. Am., 81(5), 430

1754-1782, 1991. 431

Castro, R. R., Anderson, J. G., and Singh, S. K.: Site response, attenuation and source spectra of S waves along 432

the Guerrero, Mexico, subduction zone, Bull. Seismol. Soc. Am., 80(6A), 1481-1503, 1990. 433

Castro, R. R., Fabriol, H., Bour, M., and Le Brun, B.: Attenuation and site effects in the region of Guadeloupe, 434

Lesser Antilles, Bull. Seismol. Soc. Am., 93(2), 612-626, 2003. 435

Castro, R. R., Pacor, F., Sala, A., and Petrungaro, C.: S wave attenuation and site effects in the region of Friuli, 436

Italy, J. Geophys. Res. B: Solid Earth, 101(B10), 22355-22369, 1996. 437

Chung, T. W., and Sato, H.: Attenuation of high-frequency P and S waves in the crust of southeastern South 438

Korea, Bull. Seismol. Soc. Am., 91(6), 1867-1874, 2001. 439

De Lorenzo, S., Del Pezzo, E., and Bianco, F.: Qc, Qβ, Qi and Qs attenuation parameters in the Umbria–Marche 440

(Italy) region, Phys. Earth Planet. Inter. 218, 19-30, 2013. 441

Dutta, U., Biswas, N., Martirosyan, A., Papageorgiou, A., and Kinoshita, S.: Estimation of earthquake source 442

parameters and site response in Anchorage, Alaska from strong-motion network data using generalized 443

inversion method, Phys. Earth Planet. Inter., 137(1-4), 13-29, 2003. 444

Geological Survey of India (GSI) (1998) Geological maps of India. Geological Survey of India, Kolkata 445

Harinarayan, N. H., and Kumar, A.: Site classification of the strong motion stations of Uttarakhand, India, 446

Based on the model horizontal to vertical spectral ratio, Geotechnical Frontiers, 141-149, 2017 a. 447

Harinarayan, N. H., and Kumar, A.: Seismic Site Classification of Recording Stations in Tarai Region of 448

Uttarakhand, from Multiple Approaches, Geotech. Geol. Eng., 1-16, 2017 b. 449

Harinarayan, N. H., and Kumar, A.: Determination of NEHRP Site Class of Seismic Recording Stations in the 450

Northwest Himalayas and Its Adjoining Area Using HVSR Method, Pure Appl. Geophys., 175(1), 89-107, 451

2018. 452

Hartzell, S. H.: Site response estimation from earthquake data, Bull. Seismol. Soc. Am., 82(6), 2308-2327. 453

1992. 454

Hassani, B., Zafarani, H., Farjoodi, J., and Ansari, A.: Estimation of site amplification, attenuation and source 455

spectra of S-waves in the East-Central Iran, Soil Dyn. Earthquake Eng., 31(10), 1397-1413, 2011. 456

IS 1893: Part 1–2016. Indian standard criteria for earthquake resistant design of structures—part 1: General 457 Provisions and Buildings, Bureau of Indian Standards, New Delhi, India. 458

Iwata, T., Irikura, K.: Source parameters of the 1983 Japan Sea earthquake sequence, J. of Physics of the Earth, 459 36(4), 155-184, 1988. 460

Joshi, A., Mohanty, M., Bansal, A. R., Dimri, V. P., and Chadha, R. K.: Use of spectral acceleration data for 461

determination of three-dimensional attenuation structure in the Pithoragarh region of Kumaon Himalaya, J. 462

Seismolog., 14(2), 247-272, 2010. 463

Joshi, A.: Use of acceleration spectra for determining the frequency-dependent attenuation coefficient and 464

source parameters, Bull. Seismol. Soc. Am., 96(6), 2165-2180, 2006. 465

Kim, K. D., Chung, T. W., and Kyung, J. B.: Attenuation of high-frequency P and S waves in the crust of 466

Choongchung provinces, central South Korea, Bull. Seismol. Soc. Am., 94(3), 1070-1078, 2004. 467

Konno, K., and Ohmachi, T.: Ground-motion characteristics estimated from spectral ratio between horizontal 468

and vertical components of microtremor, Bull. Seismol. Soc. Am., 88(1), 228-241, 1998. 469

Kumar, A., Mittal, H., Sachdeva, R., and Kumar, A.: Indian strong motion instrumentation network, Seismol. 470

Res. Lett., 83(1), 59-66, 2012. 471

Kumar, N., Sharma, J., Arora, B. R., and Mukhopadhyay, S.: Seismotectonic model of the Kangra–Chamba 472

sector of northwest Himalaya: Constraints from joint hypocenter determination and focal mechanism, Bull. 473

Seismol. Soc. Am., 99(1), 95-109, 2009. 474

Kvamme, L. B., and Havskov, J.: Q in southern Norway, Bull. Seismol. Soc. Am., 79(5), 1575-1588, 1989. 475

Lermo, J., and Chávez-García, F. J.: Site effect evaluation using spectral ratios with only one station. Bull. 476

Seismol. Soc. Am., 83(5), 1574-1594, 1993. 477

Luzi, L., Puglia, R., Pacor, F., Gallipoli, M.R., Bindi D, and Mucciarelli, M.: Proposal for a soil classification 478

based on parameters alternative or complementary to Vs30, Bull. Seismol. Soc. Am., 9(6), 1877-1898, 2011. 479

Ma‘hood, M., Hamzehloo, H., and Doloei, G. J.: Attenuation of high frequency P and S waves in the crust of the 480

East-Central Iran, Geophys. J. Int., 179(3), 1669-1678, 2009. 481

Mahajan, A.K. and Kumar, S.K.: Macroseismic field observations of January 26th, 2001 Kachchh earthquake 482

and its seismotectonics, J. Asian Earth Sci., 23: 17–23, 2004. 483

Menke, W.: Geophysical data analysis: discrete inverse theory: MATLAB edition (Vol. 45). Academic press, 484

2012. 485

Mukhopadhyay, S., and Kayal, J. R.: Seismic tomography structure of the 1999 Chamoli earthquake source area 486

in the Garhwal Himalaya, Bull. Seismol. Soc. Am., 93(4), 1854-1861, 2003. 487

Mukhopadhyay, S., Sharma, J., Del-Pezzo, E., and Kumar, N.: Study of attenuation mechanism for Garwhal–488

Kumaun Himalayas from analysis of coda of local earthquakes, Phys. Earth Planet. Inter., 180(1-2), 7-15, 2010. 489

Nakamura, Y.: A method for dynamic characteristics estimation of subsurface using microtremor on the ground 490

surface, Railway Technical Research Institute, Quarterly Reports, 30(1), 1989. 491

Nath, S. K., Sengupta, P., and Kayal, J. R.: Determination of S-wave site response in the Garhwal Himalaya 492

from the aftershock sequence of the 1999 Chamoli earthquake, Bull. Seismol. Soc. Am., 92(3), 1072-1081, 493

2002. 494

Negi, S. S., Paul, A., and Joshi, A.: Body wave crustal attenuation characteristics in the Garhwal Himalaya, 495

India, Pure Appl. Geophys., 172(6), 1451-1469, 2015. 496

Oth, A., Bindi, D., Parolai, S., and Di, Giacomo, D.: Spectral Analysis of K-NET and KiK-net Data in Japan, 497

Part II: On Attenuation Characteristics, Source Spectra, and Site Response of Borehole and Surface 498

StationsSpectral Analysis of K-NET and KiK-net Data in Japan, Part II, Bull. Seismol. Soc. Am., 101(2), 667-499

687, 2011. 500

Oth, A., Bindi, D., Parolai, S., and Wenzel, F.: S-wave attenuation characteristics beneath the Vrancea region in 501

Romania: new insights from the inversion of ground-motion spectra, Bull. Seismol. Soc. Am., 98(5), 2482-502

2497, 2008. 503

Oth, A., Parolai, S., Bindi, D., and Wenzel, F.: Source spectra and site response from S waves of intermediate-504

depth Vrancea, Romania, earthquakes, Bull. Seismol. Soc. Am., 99(1), 235-254, 2009. 505

Pandey, B., Jakka, R. S., and Kumar, A.: Influence of local site conditions on strong ground motion 506

characteristics at Tarai region of Uttarakhand, India, Nat. Hazards, 81(2), 1073-1089, 2016 a. 507

Pandey, B., Jakka, R. S., Kumar, A., and Mittal, H.: Site Characterization of Strong‐ Motion Recording Stations 508

of Delhi Using Joint Inversion of Phase Velocity Dispersion and H/V Curve. Bull. Seismol. Soc. Am., 106(3), 509

1254-1266, 2016 b. 510

Parolai, S., Bindi, D., Baumbach, M., Grosser, H., Milkereit, C., Karakisa, S., and Zünbül, S.: Comparison of 511

different site response estimation techniques using aftershocks of the 1999 Izmit earthquake, Bull. Seismol. Soc. 512

Am., 94(3), 1096-1108, 2004. 513

Penrose, R.: A Generalized Inverse for Matrices, Proc. Cambridge Phil. Soc. 51, 406-413, 1955. 514

Philip, G., Suresh, N., and Bhakuni, S.S.: Active tectonics in the northwestern outer Himalaya: evidence of 515

large-magnitude palaeoearthquakes in Pinjaur Dun and the Frontal Himalaya, Current Science, 106, 211–222, 516

2014. 517

Saikia, S., Chopra, S., Baruah, S., Baidya, P. R., and Singh, U.K.: Crustal imaging of the Northwest Himalaya 518

and its foredeep region from teleseismic events, Geomat Nat. Haz. Risk Journal, 7(4), 1265-1286, 2016. 519

Sarkar, I., Pachauri, A.K. and Israil, M.: On the damage caused by the Chamoli earthquake of 29 March, 1999, 520

J. Asian Earth Sci., 19: 129–134, 2001. 521

Sharma, B., Chingtham, P., Sutar, A. K., Chopra, S., and Shukla, H.P.: Frequency dependent attenuation of 522

seismic waves for Delhi and surrounding area, India, Ann. Geophys., 58(2), 0216, 2015. 523

Sharma, J., Chopra, S., and Roy, K. S.: Estimation of source parameters, quality factor (QS), and site 524

characteristics using accelerograms: Uttarakhand Himalaya region, Bull. Seismol. Soc. Am., 104(1), 360-380, 525

2013. 526

Shoji, Y., and Kamiyama, M.: Estimation of local site effects by a generalized inversion scheme using observed 527

records of ‗Small-Titan‘, Soil Dyn. Earthq. Eng., 22(9), 855-864, 2002. 528

Singh, C., Singh, A., Bharathi, V. S., Bansal, A. R., and Chadha, R. K.: Frequency-dependent body wave 529

attenuation characteristics in the Kumaun Himalaya, Tectonophysics, 524, 37-42, 2012. 530

Srivastava, H.N., Bansal, B.K., and Verma, M.: Largest earthquake in Himalaya: An appraisal, J. Geol. Soc. 531

India, 82, 15–22, 2013. 532

Tripathi, J. N., Singh, P., and Sharma, M. L.: Attenuation of high-frequency P and S waves in Garhwal 533

Himalaya, India, Tectonophysics, 636, 216-227, 2014. 534

Valdiya, K.S.: Aspects of Tectonics, Focus on south-central Asia, Tata McGraw-Hill Publishing Company Ltd., 535

New Delhi, 319, 1984. 536

Yaghmaei-Sabegh, S.: Characteristics of near-source ground motions from the 2012 Varzaghan-Ahar double 537

Earthquakes, Northwest of Iran, Nat. Hazards, 70, 1077–1097, 2014. 538

Yoshimoto, K., Sato, H., and Ohtake, M.: Frequency-dependent attenuation of P and S waves in the Kanto area, 539

Japan, based on the coda-normalization method, Geophys. J. Int., 114(1), 165-174, 1993. 540

541

Appendix A 542

In the matrix form, following the notations of Menke (1989) and incorporating the weighting factors 543

and , Eq. (1) can be written in accordance with Castro et al., (1990) as: 544

(A) (X) (b) 545

(A1) 546

The hypocentral distances of the data set is discretized into number of bins of equal widths and the value of 547

( ) is computed at each bin. The width of the bins are selected such that there is almost equal number of 548

data points in every bin. Further, ( ) versus hypocentral distance curves at each of the selected 549

frequencies are computed solving Eq. (2) in a least square sense, using singular value decomposition method 550

(Menke, 1989). 551

552

553

1 0 0 . …. 1 0 0 . ….

ln A1

ln U11

0 1 0 . …. 1 0 0 . ….

.

.

. . . . …. . . . . ….

.

.

1 0 0 . …. 0 1 0 . ….

. = ln Uij

. . . . . 0 1 0 . ….

ln A10

.

. . . . . . . . . . .

ln M1 0 . . . . . . . . . .

w1 0 0 . …. . . . . ….

.

0

- w 2/2 w 2 - w 2/2 . …. . . . . ….

.

0

0 - w 2/2 w 2 - w2/2 …. . . . . ….

.

0

. . . . …. . . . . ….

ln M N

.

FIGURES 554

555

Figure 1: Map of the region under study with EQs (stars), recording stations (triangles), and paths (solid-lines). 556

557

Figure 2: Distribution of hypocentral distances in the data set. 558

559

Figure 3: S wave spectral attenuation versus hypocentral distance. Note that ln A(f,R0) at reference distance is zero. 560

29

18

2021

2223

21

1817

13

9

2

01

3

0

5

10

15

20

25

30

35

Num

ber

of

reco

rds

Hypocentral distance (km)15 25 35 45 55 65 75 85 95 105 115 125 135 145 155

-3

-2.5

-2

-1.5

-1

-0.5

0

15 30 45 60 75 90 105

ln A

(f,R

)

R (km)

0.5Hz 1Hz

1.75Hz 2.5Hz

3.125Hz 4.5Hz

5.5Hz 6.25Hz

8Hz 10Hz

11.76Hz 12.5Hz

13.33Hz 14.28Hz

15Hz

561

Figure 4: Frequency dependence of the quality factor Q for hypocentral distances between 15 km to 105 km 562

563

Figure 5: Comparison of QS values of North West Himalaya with those obtained from parts of North West Himalaya 564 and Delhi region. The compared relations for Qs versus frequency are as follows: Garhwal-Kumouan Himalaya: 565

(Mukhopadhyay et al., 2010); Kinnaur Himalaya: (Kumar et al., 2014) ; Garhwal 566 Himalaya: (Negi et al., 2015); Delhi and NCR region: (Sharma et al., 2015). 567

568

Qs= 105 f 0.94

R² = 0.97

0

200

400

600

800

1000

1200

1400

1600

0 5 10 15

Q

frequency (Hz)

0

400

800

1200

1600

2000

0.5 5.5 10.5 15.5

Qs

frequency (Hz)

Garhwal-KumouanHimalaya(Mukhopadhyay etal., 2010)

Kinnaur Himalaya(Kumar et al., 2014)

Garhwal Himalaya(Negi et al., 2015)

Delhi and NCRregion (Sharma et al.,2015)

North WestHimalaya-This Study

569

Figure 6: Comparison of QS values of this study with regions of different tectonic settings of the world. 570

571

572

1

1001

2001

3001

0.5 5.5 10.5 15.5

Qs

frequency(Hz)

North West Himalaya-This study

Kato region-Japan(Yoshimoto et al.,1995)East central Iran

(Ma‘hood et al., 2009)

North East India(Padhy and Subhadra2010)Mainland Gujarathregion (Chopra et al.,2011)Egypt (Abdel 2009)

Koyna Region India(Sharma at al., 2007)

Umbria Marche-Italy(Lorenzo et al., 2013)

South Central Alaska(Dutta et al., 2004)

0

0.04

0.08

0.12

0.15 1.5 15

SA

(g)

Frequency (Hz)

(a) ANC

GINV H

GINV V

573

574

0

0.3

0.6

0.15 1.5 15

SA

(g)

Frequency (Hz)

(b) IGN

GINV H

GINV V

0

0.2

0.4

0.15 1.5 15

SA

(g)

Frequency (Hz)

(c) JAMI

GINV H

GINV V

575

576

0

0.03

0.06

0.09

0.15 1.5 15

SA

(g)

Frequency (Hz)

(d) BAR

GINV H

GINV V

0

0.05

0.1

0.15

0.15 1.5 15

SA

(g)

Frequency (Hz)

(e) GHA

GINV H

GINV V

577 578

Figure 7 (a-f). Site amplitude curves obtained using GINV for horizontal component and vertical component 579

580

0

0.1

0.2

0.3

0.15 1.5 15

SA

(g)

Frequency (Hz)

(f) SND

GINV H

GINV V

0

1

2

3

4

5

6

7

0.15 1.5 15

H/V

Frequency (Hz)

(a) ALIP

HVSR

GINV

581

582

0

0.5

1

1.5

2

2.5

3

3.5

4

0.15 1.5 15

H/V

Frequency (Hz)

(b) IGI

HVSR

GINV

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0.15 1.5 15

H/V

Frequency (Hz)

(c) IIT

HVSR

GINV

583

584

0

0.5

1

1.5

2

2.5

3

3.5

0.15 1.5 15

H/V

Frequency (Hz)

(d) PALHVSR

GINV

0

1

2

3

4

5

6

0.15 1.5 15

H/V

Frequency (Hz)

(e) BAL

HVSR

GINV

585

586

0

1

2

3

4

5

6

0.15 1.5 15

H/V

Frequency (Hz)

(f) RAM

HVSR

GINV

0

1

2

3

4

5

6

0.15 1.5 15

H/V

Frequency (Hz)

(g) SAL

HVSR

GINV

587

588

Figure 8 (a-i): Horizontal to vertical ratio curve obtained using GINV and HVSR method. 589

590

591

592

593

594

595

0

2

4

6

8

10

12

14

16

18

0.15 1.5 15

H/V

Frequency (Hz)

(h) MUK

GINV

HVSR

0

1

2

3

4

5

6

7

8

9

0.15 1.5 15

H/V

Frequency (Hz)

(i) PTI

HVSR

GINV

596

597

Figure 9: Vs30 as a function of fpeak (a) and Apeak (b) of GINV method for recording stations at Delhi and Tarai region 598 of Uttarakhand. 599

600

601

602

603 604

605

log (Vs30 )= 0.48 log (fpeak)+ 2.3

R² = 0.71

2.3

2.4

2.5

2.6

2.7

2.8

0 0.2 0.4 0.6 0.8

log (

Vs3

0)

(m/s

)

log (fpeak)

(a)

log (Vs30) = -0.74 log (Apeak) + 2.93

R² = 0.482.2

2.3

2.4

2.5

2.6

2.7

2.8

0.2 0.4 0.6 0.8

log (

Vs3

0)

(m/s

)

log (Apeak)

(b)

TABLES 606

Table 1: Detail of strong motion recording stations. 607

Si.

no

Station

Code

GINV HVSR

R

* or

S#

Si.

no

Station

Code

GINV HVSR

R

* or

S#

Lat.(°

) (N)

Lon. (°)

(E) fpeak

Apea

k fpeak

Apea

k

Lat.(°

) (N)

Lon. (°)

(E)

fpea

k

Apea

k

fpea

k

Apea

k

-1 -2 -3 -4 -5 -6 -7 -8 -9 -1 -2 -3 -4 -5 -6 -7 -8 -9

Himachal Pradesh

Uttarakhand

1 AMB 31.7 76.1 1.7 4.3 1.2 9.3 S 1 ALM 29.6 79.7 2.1 3.0 2.8 4.4 S

2 BHA 31.6 77.9 4.5 4.0 4.1 4.4 S 2 BAG 29.8 79.8 1.5 4.7 1.5 5.2 S

3 CHM 30.4 79.3 1.4 5.4 1.5 7.5 S 3 BAR 30.8 78.2 3.0 4.5 2.8 7.0 S

4 DEH 31.9 76.2 6.8 3.5 10 5.4 R 4 CHM 32.6 76.1 3.6 2.4 2.0 2.9 S

5 DHH 32.2 76.3 2.7 4.5 2.7 4.9 S 5 CHP 29.3 80.1 5.4 5.2 5.6 6.5 S

6 HAM 31.7 76.5 2.9 3.3 3.1 6.6 S 6 CKR 30.7 77.9 2.1 3.8 2.0 4.5 S

7 JUB 31.1 77.7 5.8 3.3 5.6 4.9 S 7

CMB

B 30.0 79.5 8.3 2.7 8.3 6.3 R

8 KLG 32.6 77.0 8.0 1.8 8.3 2.2 R 8 DHA 29.8 80.5 3.1 3.3 2.7 5.5 S

9 KUL 32.0 77.1 3.3 2.3 3.1 3.0 S 9 DNL 30.4 78.2 2.8 3.3 2.0 7.1 S

10 MAN 31.7 76.9 2.5 3.4 2.3 5.3 S 10 DUN 30.3 78.0 2.9 5.8 3.1 7.1 S

11 RAM 31.4 77.6 2.8 4.2 2.7 5.6 S 11 GAR 30.1 79.3 2.4 3.4 2.3 4.5 S

12 SAL 32.7 76.1 2.7 3.0 2.7 5.8 S 12 GHA 30.4 78.7 5.2 2.3 4.5 5.5 S

13 SND 31.5 76.9 5.0 3.5 5.0 4.2 S 13 GLTR 30.3 79.1 2.9 3.5 2.8 6.1 S

14 SOL 30.9 77.1 0.5 2.6 0.6 4.6 S 14 JSH 30.5 79.6 1.4 2.2 1.5 3.0 S

15 UNA 31.5 76.3 1.5 3.9 1.8 6.0 S 15 KAP 29.9 79.9 3.7 6.4 3.3 9.2 S

16 KJK 30.9 76.9 0.7 6.4 0.4 12.0 S 16 KHA 28.9 80.0 1.3 4.2 2.0 8.0 S

17 PLM 32.1 76.5 1.7 1.8 1.6 2.2 S 17 KKHR 30.2 78.9 9.5 3.5 9.5 9.3 R

Punjab

18 KOT 29.7 78.5 0.7 2.4 0.7 3.3 S

1 ANS 31.2 76.5 0.9 4.9 0.9 17.0 S 19 KSK 29.2 79.0 3.1 3.8 3.2 9.9 S

2 ASR 31.6 74.9 1.0 2.9 1.0 8.0 S 20 KSL 30.9 77.0 3.0 3.9 2.1 19.4 S

3 GSK 31.2 76.1 0.8 2.3 0.8 3.8 S 21 LANG 30.3 79.3 7.7 2.8 7.9 5.7 R

4 JAL 31.3 75.6 2.9 1.5 2.9 1.7 S 22 LAN 29.8 78.7 1.4 2.8 1.4 6.1 S

5 KAT 31.4 75.4 2.0 3.1 3.0 3.4 S 23 MUN 30.1 80.2 4.3 2.8 7.0 3.6 S

6 MOG 30.8 75.2 2.2 2.3 2.2 3.9 S 24 PAU 30.2 78.8 5.9 2.1 3.1 3.7 S

7 MUK 31.9 75.6 1.4 4.7 1.4 15.9 S 25 PTH 29.6 80.2 8.0 2.7 4.6 3.1 R

8 NAW 31.1 76.1 1.4 3.6 1.4 3.5 S 26 PTI 29.4 79.9 4.0 6.6 3.6 8.0 S

9 NKD 31.1 75.5 1.2 2.2 1.3 3.4 S 27 RIS 30.1 78.3 3.8 3.0 3.4 6.4 S

10 PHG 31.2 75.8 2.7 2.7 2.7 5.2 S 28 ROO 29.9 77.9 1.2 4.4 1.3 5.2 S

11 TAR 31.4 74.9 2.6 2.7 2.7 4.6 S 29 RUD 30.3 79.0 1.3 2.8 1.5 4.2 S

Delhi

30 SMLI 30.2 79.3 9.1 3.3 8.7 6.1 R

1 ARI 26.1 77.5 2.7 3.2 2.5 7.0 S 31 TAN 29.1 80.1 5.4 3.4 5.0 6.3 S

2 IGN 28.5 77.2 3.6 2.6 4.5 3.9 S 32 THE 30.4 78.4 1.6 2.8 1.5 3.6 S

3 JNU 28.5 77.2 9.0 2.1 8.7 3.5 R 33 UDH 29.0 79.4 2.7 6.9 2.2 10.1 S

4 DJB 28.7 77.2 2.2 3.2 10 4.5 S 34 UTK 30.7 78.4 2.4 2.9 2.3 4.4 S

5 NDI 28.7 77.2 6.8 2.5 7.2 3.7 R 35 VIK 30.5 77.8 2.3 3.8 2.3 10.4 S

6 IMD 28.7 77.2 6.0 2.0 6.3 2.9 S 36 GDRI 30.2 78.7 6.0 3.4 5.1 4.8 S

7 NTPC 28.5 77.3 2.8 3.6 2.8 5.4 S 37 TLWR 30.3 79.0 1.1 2.1 1.0 4.6 S

8 ANC 28.5 77.3 4.6 3.2 4.5 4.5 S 38 UKMB 30.3 79.1 1.0 2.9 1.4 10.0 S

9 JAMI 28.6 77.3 4.7 3.2 4.5 7.3 S 39 ADIB 30.2 79.2 6.4 12.7 6.3 17.8 R

10 LDR 28.6 77.2 0.7 4.3 0.9 7.0 S 40 NUTY 30.2 79.2 4.8 2.4 4.7 4.3 S

11 VCD 28.6 77.2 4.6 2.7 4.6 3.6 S 41 KHIB 30.2 78.8 7.7 3.3 8.0 8.7 R

12 IIT 28.6 77.3 4.3 2.9 4.5 4.3 S 42 STRK 30.3 79.0 4.7 2.8 4.7 5.8 S

13 NSIT 28.6 77.0 2.4 2.5 2.3 3.9 S 43 NANP 30.3 79.3 3.9 3.5 3.8 9.1 S

14 RGD 28.7 77.1 2.3 2.1 2.9 3.8 S

Haryana

15 GGI 28.7 77.2 15 5.3 15 8.4 R 1 PAL 28.1 77.3 2.8 2.7 2.9 3.4 S

16 DLU 28.7 77.2 1.8 3.2 1.9 3.7 S 2 JAFR 28.6 76.9 6.0 2.0 7.1 2.6 S

17 DCE 28.8 77.1 3.8 3.1 4.7 4.2 S 3 GUR 28.4 77.0 1.0 4.1 1.0 5.2 S

18 IGI 28.6 77.1 2.2 2.4 2.2 3.8 S 4 REW 28.2 76.6 2.5 2.1 2.5 3.7 S

19 ZAKI 28.6 77.2 3.9 3.5 3.9 8.4 S 5 SON 29.0 77.0 1.0 3.5 2.8 4.0 S

20 ALIP 28.8 77.1 2.3 3.2 2.5 6.9 S 6 ROH 28.6 77.2 1.4 3.1 2.0 4.6 S

21 ROI 28.6 77.2 1.4 3.1 2.0 4.6 S 7 CRRI 29.0 77.1 4.3 3.5 4.4 9.3 S

R* Rock site

8 BAL 28.3 77.3 1.5 3.0 1.4 5.8 S

S# Soil site

9 KAI 29.8 76.4 1.2 3.0 1.2 6.5 S

608

Table 2: Details of earthquake events (from PESMOS) considered for estimation of site parameters in 609 this work. 610

Eve

nt

No.

dd/mm/yy

yy

Lat

.

Lon

g.

Dept

h

Magnitu

de

Eve

nt

No.

dd/mm/yy

yy

Lat

.

Lon

g.

Dept

h

Magnitu

de

-1 -6 -2 -3 -4 -5

-1 -6 -2 -3 -4 -5

1

14-12-

2005

30.

9 79.3 25.7 5.2

44

24-09-

2011

30.

9 78.3 10.0 3.0

2

07-05-

2006

28.

7 76.6 20.2 4.1

45

26-10-

2011

31.

5 76.8 5.0 3.5

3

29-11-

2006

27.

6 76.7 13.0 3.9

46

16-01-

2012

29.

7 78.9 10.0 3.6

4

10-12-

2006

31.

5 76.7 33.0 3.5

47

12-03-

2012

28.

9 77.3 5.0 3.5

5

22-07-

2007

29.

9 77.9 33.0 5.0

48

26-02-

2012

29.

6 80.8 10.0 4.3

6

25-11-

2007

28.

6 77.0 20.3 4.3

49

27-03-

2012

26.

1 87.8 12.0 3.5

7

04-10-

2007

32.

5 76.0 10.0 3.8

50

05-03-

2012

28.

7 76.6 14.0 4.9

8

18-10-

2007

28.

3 77.6 5.6 3.6

51

28-07-

2012

29.

7 80.7 10.0 4.5

9

19-08-

2008

30.

1 80.1 15.0 4.3

52

23-08-

2012

28.

4 82.7 5.0 5.0

10

19-10-

2008

29.

1 76.9 7.0 3.2

53

02-10-

2012

32.

4 76.4 10.0 4.9

11

21-10-

2008

31.

5 77.3 10.0 4.5

54

03-10-

2012

32.

4 76.3 10.0 3.6

12

31-01-

2009

32.

5 75.9 10.0 3.7

55

06-11-

2012

32.

3 76.2 5.0 4.1

13

09-01-

2009

31.

7 78.3 16.0 3.8

56

11-11-

2012

29.

3 80.1 5.0 5.0

14

25-02-

2009

30.

6 79.3 10.0 3.7

57

15-11-

2012

30.

2 80.1 5.0 3.0

15

18-03-

2009

30.

9 78.2 10.0 3.3

58

27-11-

2012

30.

9 78.4 12.0 4.8

16

04-09-

2008

30.

1 80.4 10.0 5.1

59

19-12-

2012

28.

6 76.8 10.0 2.9

17 01-05- 29. 80.1 10.0 4.6

60 02-01- 29. 81.1 10.0 4.8

2009 9 2013 4

18

15-05-

2009

30.

5 79.3 15.0 4.1

61

09-01-

2013

29.

8 81.7 5.0 5.0

19

17-07-

2009

32.

3 76.1 39.3 3.7

62

10-01-

2013

30.

1 80.4 5.0 3.2

20

27-08-

2009

30.

0 80.0 14.0 3.9

63

29-01-

2013

30.

0 81.6 7.0 4.0

21

21-09-

2009

30.

9 79.1 13.0 4.7

64

11-02-

2013

31.

0 78.4 5.0 4.3

22

03-10-

2009

30.

0 79.9 15.0 4.3

65

17-02-

2013

30.

9 78.4 10.0 3.2

23

06-12-

2009

35.

8 77.3 60.0 5.3

66

01-05-

2013

33.

1 75.8 15.0 5.8

24

11-01-

2010

29.

7 80.0 15.0 3.9

67

05-09-

2013

30.

9 78.5 11.0 3.5

25

22-02-

2010

30.

0 80.1 2.0 4.7

68

11-11-

2013

28.

5 77.2 10.0 3.1

26

24-02-

2010

28.

6 76.9 17.0 2.5

69

11-11-

2013

28.

4 77.2 11.0 2.8

27

14-03-

2010

31.

7 76.1 29.0 4.6

70

11-11-

2013

28.

4 77.2 12.0 2.5

28

03-05-

2010

30.

4 78.4 8.0 3.5

71

11-11-

2013

28.

4 77.2 13.0 3.1

29

28-05-

2010

31.

2 77.9 43.0 4.8

72

16-04-

2013

28.

0 62.1 16.0 7.8

30

31-05-

2010

30.

0 79.8 10.0 3.6

73

04-06-

2013

32.

7 76.7 18.0 4.8

31

06-07-

2010

29.

8 80.4 10.0 5.1

74

05-06-

2013

32.

8 76.3 10.0 4.5

32

10-07-

2010

29.

9 79.6 10.0 4.1

75

09-07-

2013

32.

9 78.4 10.0 5.1

33

26-01-

2011

29.

0 77.2 10.0 3.2

76

13-07-

2013

32.

2 76.3 10.0 10.0

34

14-03-

2011

30.

5 79.1 8.0 3.3

77

15-07-

2013

32.

6 76.7 30.0 4.4

35

18-02-

2011

28.

6 77.3 5.0 2.3

78

02-08-

2013

33.

5 75.5 20.0 5.4

36

09-02-

2011

30.

9 78.2 10.0 5.0

79

29-08-

2013

31.

4 76.1 10.0 4.7

37

04-04-

2011

29.

6 80.8 10.0 5.7

80

20-10-

2013

35.

8 77.5 80.0 5.5

38

15-06-

2011

30.

6 80.1 10.0 3.6

81

25-12-

2013

31.

2 78.3 10.0 4.0

39

20-06-

2011

30.

5 79.4 12.0 4.6

82

17-06-

2014

32.

2 76.1 10.0 4.1

40

23-06-

2011

30.

0 80.5 5.0 3.2

83

21-08-

2014

32.

3 76.5 10.0 5.0

41

28-07-

2011

33.

3 76.0 21.0 4.4

84

29-11-

2015

30.

6 79.6 15.0 4.0

42

07-09-

2011

28.

6 77.0 8.0 4.2

85

25-09-

2016

30.

0 79.5 11.0 3.7

43

21-09-

2011

30.

9 78.3 10.0 3.1

86

01-12-

2016

30.

6 79.6 19.0 4.0

611

Table 3: List of Earthquakes and the corresponding stations considered for the estimation of path 612 parameter. 613

Earthquake Event Stations

25-11- 2007 HGR, NDI, CRRI, PAL, REW, NDI , CRRI, LDR, JAFR, IIT

19-08-2008 CHP, PTH, KAP, MUN

04-09- 2008 MUN, CHP, PTH, DHA , KAP, GHA, JSH

01-05-2009 MUN, BAG, KAP, GAR, CHM

17-07-2009 DHA, KLG

27-08-2009 KAP, BAG, MUN

03-08-2009 KAP, CHP, BAG

11-01-2010 PTH, CHP, DHA

22-02-2010 KAP, BAG, DHA, ROO, UDH

24-02-2010 RGD, IGN, ROH, DJB, CHP, ANC, JAMI, GGI, DLU, DCE

14-03-2010 DEH, JUB, SND, BHA, HAM, GAR, JAL, KAP, AMB, ROO

03-04-2010 THE, BAR, DNL, ROO

28-05-2010 JUB, BAR, ROO, UNA

06-06-2010 MUN, CHP

10-07-2010 BAG, KAP, GAR, ROO

18-02-2011 DJB, ANC

09-02-2011 UTK, SND, KUL, CKR

04-04-2011

JSH, CHP, PTI, PTH, ALM, DDH, BAG, DHA, GAR, MUN, RUD, THE, CHM,

BAR, SND, KOT, DNL, LDR, ROO, TAN, KHA, UDH, KSH, DUD

12-03-2012 GGI, ANC, DLU, DCE

27-03-2012 ARI, ANC

05-03-2012

JAFR, JNU, DJB, IMD, PLW, GUR, NOI, NTPC, ANC, IIT, NSIT, ZAKI, ROO,

RGD, GGI, DLU, DCE, ALIP, SON, BAR, KAI, NKD,

02-10-2012 CHA, RAM

11-11-2012 CHA, CHP, PTH

27-11-2012 UTK, THE, DNL, CKR

02-01-2013 CHP, PTI, PTH

09-01-2013 CHP, PTI, PTH, TAN

11-02-2013 UTK, ROO

11-11-2013 (19:11:18) NTPC, IGN, JNU, DJB, IMD, VCD, IGI, RGD, GGI, DLU, DCE, ALIP

11-11-2013 (22,10,42) IGN, JNU, DJB, VCD, RGD, DCE, ALIP

11-11-2013 (20:11:30) IGN, JNU, DJB, VCD, RGD, GGI, DLU, DCE

29-08- 2013 GSK, RAM, ROO, NKD, ANS, KAT

25-09-2016 UKMB, CMBB, GDRI, DURD

614

Table 4: Resulting parameters of eq. 7. 615

f (Hz) m

( )

( )⁄

(1) (2) (3)

0.50 0.0095 51.65

1.00 0.0101 97.15

1.75 0.0115 149.32

2.50 0.0096 255.53

3.12 0.0089 344.54

3.57 0.0093 376.67

4.50 0.0098 450.57

5.00 0.0074 663.01

5.50 0.0089 606.39

6.25 0.0084 730.09

7.14 0.011 636.92

8.00 0.013 603.84

10.00 0.0172 570.49

11.76 0.0124 930.60

12.50 0.0102 1202.51

13.33 0.0098 1334.70

14.28 0.0115 1218.46

15.00 0.0108 1362.85

Table 5: fpeak, Apeak and Vs30 values for 27 stations located in Terai region of Uttarakhand and Delhi 616 region. 617

GINV

Station

Code fpeak,(Hz) Apeak, Vs30 (m/s)

IGN 3.6 2.6 493*

JNU 6.5 2.05 565*

DJB 5.22 3.2 543*

NDI 6.8 2.5 493*

IMD 6 2 543*

NTPC 2.8 3.6 345*

ANC 4.6 3.2 564*

JAMI 4.7 3.15 346*

LDR 0.7 4.32 270*

VCD 5.6 2.7 550*

IIT 4.3 2.94 332*

NSIT 2.4 2.5 391*

RGD 2.3 2.07 346*

DLU 1.81 3.2 323*

DCE 3.78 3.1 328*

IGI 2.2 2.4 360*

ZAKI 3 3.49 337*

ROI 2.3 3.24 303*

ALIP 1.4 3.09 338*

DUN 2.9 5.8 289**

KHA 1.3 4.2 218**

KSK 3.13 3.75 208**

RIS 3.8 3 331**

ROO 1.16 4.35 218**

TAN 5.4 3.4 434**

UDH 2.74 6.9 198**

VIK 2.29 3.8 424**

**Pandey et al., 2016a; * Pandey et al., 2016b 618

619

620

621

622

623

624

625

626

627

List of Figures 628

1. Figure 1: Map of the region under study with EQs (stars), recording stations (triangles), and paths (solid-629 lines). 630

2. Figure 2: Distribution of hypocentral distances in the data set. 631

3. Figure 3: S wave spectral attenuation versus hypocentral distance. Note that Log A(f,R0) at reference distance 632 is zero. 633

4. Figure 4: Frequency dependence of the quality factor Q for hypocentral distances between 15km to 105km 634

5. Figure 5: Comparison of QS values of North West Himalaya with those obtained from parts of North West 635 Himalaya and Delhi region. The compared relations for Qs versus frequency are as follows: Garhwal-Kumouan 636 Himalaya: (Mukhopadhyay et al., 2010); Kinnaur Himalaya: (Kumar et al., 637 2014) ; Garhwal Himalaya: (Negi et al., 2015); Delhi and NCR region: 638 (Sharma et al., 2015). 639

6. Figure 6: Comparison of QS values of this study with regions of different tectonic settings of the world. 640

7. Figure 7: Site amplitude curves obtained using GINV for horizontal component and vertical component 641

8. Figure 8: Horizontal to vertical ratio curve obtained using GINV and HVSR method 642

9. Figure 9: Vs30 as a function of fpeak (a) and Apeak (b) of GINV method for recording stations at Delhi and 643 Tarai region of Uttarakhand. 644

List of Tables 645

Table 1: Detail of strong motion recording stations. 646

Table 2: Details of earthquakes considered for estimation of site parameters in this work 647

Table 3: List of Earthquakes and the corresponding stations considered for the estimation of path parameter. 648

Table 4: Resulting parameters of Eq. (7). 649

Table 5: fpeak, Apeak and Vs30 values for 27 stations located in Terai region of Uttarakhand and Delhi region. 650

651